may 2017 rff dp 17-09 discussion paperthe cost of natural disasters has been steadily increasing...
TRANSCRIPT
May 2017 RFF DP 17-09
Defining the Roles of the Public and Private Sector in Risk Communication, Risk Reduction, and Risk Transfer
Car o l yn Kousky and How ard Kunreut her
DIS
CU
SS
ION
PA
PE
R
© 2017 Resources for the Future. All rights reserved. No portion of this paper may be reproduced without
permission of the authors.
Resources for the Future (RFF) is an independent, nonpartisan organization that conducts rigorous economic
research and analysis to help leaders make better decisions and craft smarter policies about natural resources and the
environment.
Discussion papers are research materials circulated by their authors for purposes of information and discussion.
They have not necessarily undergone formal peer review. Unless otherwise stated, interpretations and conclusions in
RFF publications are those of the authors. RFF does not take institutional positions.
Defining the Roles of the Public and Private Sector in Risk
Communication, Risk Reduction, and Risk Transfer
Carolyn Kousky and Howard Kunreuther
Abstract
Insurance is an essential component of household and community resilience: it protects insureds
financially against disaster losses, can encourage investments in cost-effective mitigation measures
through premium reductions, and facilitates rebuilding of property and long-term recovery following a
disaster via claim payments. Private insurers face challenges in providing protection against low-
probability, high-consequence events, however, and the perceived market failures have led governments
around the world to create various (quasi to fully) public insurance entities, often designed as public–
private partnerships. This paper synthesizes findings from six papers and the resulting discussion at a
November 2016 workshop, “Improving Disaster Financing: Evaluating Policy Interventions in Disaster
Insurance Markets.” Participants evaluated disaster insurance programs for flood, earthquake, and
terrorism losses, as well as comprehensive homeowners policies. This paper discusses the difficulties in
providing protection against these types of disasters and suggests ways to improve public-private
partnerships for disaster financing in three interrelated areas: (1) risk communication, (2) risk reduction,
and (3) risk transfer. The paper concludes with a proposal for a comprehensive insurance program that
could harness the benefits of both the public and private sectors.
Contents
1. Introduction ......................................................................................................................... 1
2. The Disaster Insurance Gap .............................................................................................. 2
2.1. Consumer Behavior ..................................................................................................... 3
2.2. Insurer Behavior........................................................................................................... 4
2.3. Regulator Behavior ...................................................................................................... 6
3. The Emergence of Public Private Disaster Insurance Partnerships .............................. 6
4. Roles of the Public and Private Sectors .......................................................................... 10
4.1. Risk Communication ................................................................................................. 10
4.2. Risk Reduction ........................................................................................................... 13
4.3. Risk Transfer .............................................................................................................. 16
5. A Proposed Program ........................................................................................................ 19
6. Conclusion and Suggestions for Future Research ......................................................... 21
6.1. Finer-Scale Risk Communication .............................................................................. 21
6.2. Higher Take-Up of Insurance .................................................................................... 22
6.3. Insurance for Low-Income Households ..................................................................... 22
6.4. New Policy Models .................................................................................................... 23
References .............................................................................................................................. 24
Resources for the Future Kousky and Kunreuther
1
Defining the Roles of the Public and Private Sector in Risk
Communication, Risk Reduction, and Risk Transfer
Carolyn Kousky and Howard Kunreuther
1. Introduction
The cost of natural disasters has been steadily increasing around the world largely
because more properties—and more valuable properties—are being built in risky locations (e.g.,
Benson and Clay 2004; Kousky 2014; Swiss Re Institute 2017). These rising costs have
commensurately increased governments’ liability for disaster losses. Since 2000, the US
government has spent $265 billion on supplemental appropriations for disaster relief (Lindsay
and Murray 2014). Furthermore, climate scientists predict more intense hurricanes, heavier
precipitation events, and more severe heat waves, fires, and droughts in many places across the
globe and the planet warms (e.g., IPCC 2012; Sander et al. 2013).
Insurance is an essential component of household and community resilience for reducing
future losses from floods, earthquakes, and other disasters because it encourages investments in
cost-effective loss reduction measures while facilitating the rebuilding of damaged property and
long-term recovery from disasters through claim payments (e.g., Turnham et al. 2011;
Kunreuther et al. 2013). Yet the same forces that make insurance ever more critical also increase
the challenges for underwriting these risks. Insuring disasters poses a challenge for insurance
companies because of the potential for extreme losses. To meet regulatory and rating agency
requirements, insurers must hold or purchase access to sufficient capital to remain solvent after
an event; this can be costly. The high premiums associated with insuring catastrophic risks may
lead consumers to take their chances by forgoing coverage. The severe loss of capital following a
catastrophic disaster may also lead to “hard” insurance markets, where supply is scarce and
coverage costly. These challenges have prompted various kinds of government intervention in
almost all disaster insurance markets, both in the United States and around the world.
Kousky: [email protected]; Kunreuther: [email protected].
This paper was prepared for the “Improving Disaster Financing: Evaluating Policy Interventions in Disaster
Insurance Markets” workshop held at Resources for the Future on November 29–30, 2016. We would like to thank
our sponsors of this project: the American Academy of Actuaries; the American Risk and Insurance Association;
Risk Management Solutions; the Society of Actuaries; and XL Catlin.
Resources for the Future Kousky and Kunreuther
2
This paper synthesizes findings from the dialogue at a workshop hosted by Resources for
the Future and the Wharton Risk Management and Decision Processes Center in November
2016, as well as the six papers prepared for the workshop. The papers evaluated the following
programs: the National Flood Insurance Program (NFIP), the California Earthquake Authority
(CEA), wind pools in the state of Florida, the Terrorism Risk Insurance Program (TRIA), Flood
Re in the UK, as well as all-hazards disaster insurance.
Here, we tease out lessons learned from these programs to suggest ways of improving
public-private partnerships for disaster insurance for households and small businesses in three
interrelated areas: (1) risk communication, (2) risk reduction, and (3) risk transfer. We do not
consider insurance for infrastructure or large commercial entities. We first discuss the challenges
of insuring disasters, and the nature of the decision processes of interested parties in dealing with
low-probability, high-consequence events. With this background, we formulate guiding
principles for a comprehensive insurance program and discuss the limitations of the private
sector in marketing insurance without public sector involvement.
2. The Disaster Insurance Gap
Insurance against disaster losses allows policyholders to finance repair and rebuilding
without having to divert current income or draw down savings. Insurers that give premium
discounts for hazard mitigation encourage those at risk to invest in protective measures prior to a
disaster. Insurance can also reduce disaster assistance from the federal government, which may
be slow to arrive, mismatched to meet immediate needs, and distortionary—that is, it may
generate perverse incentives (Talbot and Barder 2016). For instance, disaster aid may lead those
at risk to underinvest in risk reduction or may skew decisions toward what is funded even if that
is not optimal. Insurance, on the other hand, when priced appropriately, tends to have fewer
incentive problems.
Despite those benefits of insurance, a large disaster insurance gap exists worldwide and
in the United States (Swiss Re Institute 2017). For example, just over 10 percent of homes in
California have earthquake insurance (Marshall 2017). In 100-year floodplains of the United
States, only roughly half of homes are insured against floods, and even fewer outside these areas
carry flood insurance (e.g., Dixon et al. 2006; New York City 2013; Adams 2015). A number of
factors may suppress consumer demand for disaster coverage, impede insurers from supplying
coverage, and/or present regulatory challenges for insurance commissioners.
Resources for the Future Kousky and Kunreuther
3
Empirical studies guided by research in cognitive psychology and behavioral
decisionmaking over the past 50 years have revealed that individuals and organizations often
make suboptimal decisions about low-probability, high-consequence risks. Decisions about these
risks are often governed by combining intuitive with deliberative thinking. In his thought-
provoking book Thinking, Fast and Slow, Nobel Laureate Daniel Kahneman (2011) summarized
the differences between these two modes of thinking.
Intuitive thinking (System 1) operates automatically and quickly, with little or no effort
and no voluntary control. It is often guided by emotional reactions and simple rules of thumb
acquired by personal experience. Deliberative thinking (System 2) allocates attention to effortful
and intentional mental activities in which individuals evaluate trade-offs, recognize relevant
interdependencies, and consider the need for coordination. Most people make choices by
combining these two modes of thinking.
Intuitive processes work well when people have copious data on the outcomes of
different decisions, and when recent experience is a meaningful guide for future actions. These
processes in general do not work well for low-probability, high-consequence events, however,
because by definition, such events are rare and people have limited or no past experience with
them. For extreme events, individuals exhibit systematic biases, such as being unduly influenced
by a recent disaster, underestimating the future likelihood of a severe loss, or dismissing a risk by
treating it below their threshold level of concern.
2.1. Consumer Behavior
Low-probability, high-consequence events are subject to the availability bias, where the
judged likelihood of an event depends on its salience (Tversky and Kahneman 1973). This is a
principal reason why individuals might purchase insurance only after a disaster but cancel their
policies several years later: if they have not suffered a loss, they now perceive the likelihood of a
disaster as low. An analysis of NFIP policies revealed that the median tenure of flood insurance
was between two and four years, whereas the average length of time in a residence was seven
years (Michel-Kerjan et al. 2012). This behavior can be observed even when homeowners are
required to purchase flood insurance as a condition for obtaining a federally insured mortgage if
banks and financial institutions fail to enforce the requirement over the life of the loan.
Individuals are also myopic, in that they focus on the returns that they are likely to
receive in the next few years rather than over a longer time horizon. When determining whether
to invest in protection to reduce losses to a home or business, people often view the upfront costs
Resources for the Future Kousky and Kunreuther
4
as much higher than the expected benefits over the next two or three years, and hence they
decide not to undertake these actions. If they focused on the full life of the property, they might
realize that the expected benefits from the mitigation measure exceed the upfront costs; from a
financial perspective, then, the investment would be justified.
Other factors can cause households to be uninsured against disasters. In the United States,
standard homeowners policies exclude losses from floods and earthquakes. Homeowners may
not realize that they have to purchase separate coverage for these risks, or they may know but
decline coverage. Disaster policies may not automatically renew each year and may not be
escrowed as part of mortgage payments, leading some people to drop coverage.
Workshop participants expressed concern that federal disaster aid may create a moral
hazard problem, whereby homeowners and business owners assume the government will provide
them with assistance following damage to their property and so fail to insure. Empirical evidence
on this moral hazard is scant, however. The few papers that have tried to correlate actual disaster
spending with insurance purchases find either a positive relationship (Browne and Hoyt 2000) or
a very small effect (Kousky et al. 2014). This can be explained by the limited disaster money
given to US disaster victims and the requirement that recipients of disaster assistance maintain
insurance policies to be eligible for such aid in the future (Kousky 2016). More research is
needed on the extent to which perceptions of generous postdisaster relief may lead to lower
predisaster insurance purchases.
2.2. Insurer Behavior
Disaster losses can be very severe in some years, that is, losses are “fat tailed” (e.g.,
Malamud and Turcotte 2006; Holmes et al. 2008). In an extreme event, a large number of an
insurance company’s insured portfolio will suffer damage at the same time. To protect against
the possibility of insolvency, insurance companies are required to hold large amounts of capital
and/or secure access to capital through reinsurance or other risk transfer mechanisms; this can be
costly and increase the price of insurance significantly (Kunreuther and Michel-Kerjan 2011).
When disaster insurance is expensive, however, the premiums can be more than property owners
are willing or able to pay (Kousky and Cooke 2012).
It is somewhat surprising that sometimes insurance managers also sometimes exhibit
systematic biases or use simplified decision rules when faced with uncertainty or ambiguous
information regarding low-probability risks. Limited data and limited past experience with
Resources for the Future Kousky and Kunreuther
5
extreme events may cause insurers to engage in intuitive thinking when determining what
coverage to offer and how much to charge (Cabantous et al. 2011).
Behavior by insurers may follow a “safety-first” model, originally proposed by Roy
(1952) and applied to insurance by James Stone, the insurance commissioner of Massachusetts.
Stone (1973a, 1973b) suggested that an underwriter who wanted to determine the conditions for
a specific risk to be insurable would focus on keeping the probability of insolvency below some
threshold level of concern rather than trying to maximize expected profits. Such an approach,
Stone said, would lead insurers (1) to not charge anything for coverage against a specific risk if
its likelihood were perceived to be very low, and (2) to charge much higher premiums or view
the risk as uninsurable when the likelihood of a disaster and the resulting damage were highly
ambiguous (Hogarth and Kunreuther 1985; Kunreuther and Hogarth 1992). Discussions with
underwriters indicate that this safety-first model still characterizes insurers’ decisionmaking
process.
September 11, 2001, provides an illustration of the behavior. Before the terrorist attacks,
actuaries and underwriters had not determined a price for protection against terrorism coverage
or excluded this coverage from their standard commercial policies. They were essentially
covering this risk for the very modest add-on for unspecified events included in typical property
insurance premiums. Their failure to examine the financial risks associated with terrorism was
surprising, given the attempted bombing of the World Trade Center in 1993, the Oklahoma City
bombing in 1995, and other terrorist attacks throughout the world prior to 2001. Presumably
insurers perceived the likelihood of larger attacks on US soil to be negligible. Following 9/11,
most insurance companies changed course and refused to offer any coverage against terrorism,
considering it an uninsurable risk despite increased buyer demand. The few that did provide
insurance charged extremely high premiums (Michel-Kerjan and Kunreuther 2017).1
1 A few examples stand out. First, Golden Gate Park in San Francisco was unable to obtain terrorism coverage at
any price (Smetters 2004). Second, prior to 9/11, Chicago’s O’Hare Airport had $750 million of terrorism insurance
coverage at an annual premium of $125,000. After 9/11, insurers offered the airport only $150 million of coverage at
an annual premium of $6.9 million (Jaffee and Russell 2003). And third, in early 2002 one insurance broker
negotiated a contract between an industrial firm that was willing to pay $900,000 for $9 million in coverage should a
terrorist attack damage or destroy one of its facilities in the coming year; that is, the insurer considered the
likelihood of damage from a terrorist attack in the next year to be greater than 1 in 10 (Kunreuther et al. 2013).
Resources for the Future Kousky and Kunreuther
6
2.3. Regulator Behavior
State insurance commissioners and state legislatures authorize the sale of insurance
against different risks, oversee and approve their premiums, and regulate many other aspects of
the private insurance market. Insurance companies have voiced concerns that state regulators
limit their ability to charge prices that reflect the risks they face. It has been observed that US
insurance commissioners tend to weight the affordability and availability of insurance policies
more heavily than solvency considerations (Klein and Wang 2007). Controlling rates to appease
homeowners, however, can make it difficult for firms to operate profitably. For instance, a 2008
examination of cumulative insurer profits by year in Florida revealed that insurance companies
had been in the red since 1992 (Klein 2008). Although the companies made profits in quiet years,
the losses from 1992, 2004, and 2005 more than wiped out all those profits. This led many
insurance companies to scale back operations in Florida and along the Gulf Coast.
After an extreme event, state insurance commissioners often prevent steep price increases
in an attempt to protect consumers. Following Hurricane Andrew (1992), insurance regulators in
Florida prohibited dramatic rate increases and let companies increase prices only gradually over
a decade; the state legislature passed a moratorium on policy cancellations. The Florida
legislature also instituted a three-year moratorium that limited how quickly firms could reduce
their market share in the state (McChristian 2012). Following Hurricane Katrina (2005), Florida
allowed an initial wave of price increases, which were generally highest in coastal areas, but then
began disapproving them in 2006. Other states have taken similar measures. After Katrina, the
Louisiana Department of Insurance prohibited cancellation or nonrenewal of residential dwelling
and commercial property insurance for structures damaged by Hurricane Katrina or Rita (2005)
until 60 days after all repair and reconstruction had been completed (Klein 2008).
Workshop participants commented that companies may feel that the current regulatory
environment makes it challenging for them to develop and test new products designed to appeal
to consumers. It appears that insurance commissioners and insurance companies need to have a
more effective dialogue on how to allow for the pricing of catastrophic risk while protecting
consumers from unfair pricing practices.
3. The Emergence of Public Private Disaster Insurance Partnerships
Private insurers’ concerns about coverage against potentially catastrophic losses has led
to government interventions in these markets around the world. The creation of public sector
Resources for the Future Kousky and Kunreuther
7
programs has often been spurred by a disaster that exposed the insurance gap or created other
challenges for the private market.
Flood insurance, for example, was offered by the US private sector beginning in 1896.
Following the 1927 Mississippi floods, however, when insurers suffered serious losses, all
insurance companies withdrew coverage of property in highly flood-prone areas because of their
concern with future catastrophic losses as well as challenges with adverse selection (Kunreuther
et al. 1978). It took another disaster, flooding from Hurricane Betsy in 1965, to demonstrate how
few homes were insured against flood, prompting liberal federal disaster relief and widespread
congressional support for creating the NFIP (Knowles and Kunreuther 2014).
Similarly, Hurricane Andrew in 1992 led to the establishment of the Florida Residential
Property and Casualty Joint Underwriting Association (later Florida Citizens), the Northridge
earthquake of 1994 led to the creation of the California Earthquake Authority (CEA), the
terrorist attacks on 9/11 led to the creation of the Terrorism Risk Insurance Program, and
flooding in the United Kingdom, particularly in 1953, led to establishment of the first UK
partnership with the private sector for flood insurance, Flood Re. These quasi- to fully
governmental disaster insurance programs vary substantially in their design, spanning the range
of public-private partnership types. Some provide insurance directly to property owners, while
others offer reinsurance; some offer coverage for a single peril only, while others provide more
comprehensive policies. Take-up rates vary: some have mandatory purchase or offer
requirements, while others do not. Financing arrangements and the public-private shares for
payment of claims vary as well. The programs also differ in the nature and extent of incentives or
regulations for risk reduction and how they make coverage affordable. Table 1 compares the
programs evaluated in this project.2
2 Several other papers and reports have compared countries’ disaster insurance programs (CCS 2008; Medders et al.
2011; Paudel 2012; Atreya et al. 2015; McAneney et al. 2016) but do not draw out specific lessons for the United
States, as we do in this work. We also look beyond just the structure of risk transfer to consider public and private
responsibilities for risk reduction and risk communication as well.
Resources for the Future Kousky and Kunreuther
8
Table 1. Insurance Program Design Comparison
Coverage
provided
Take-up rates
and mandates Financing
Claims
paying
ability (up to
the given
event) a
Incentives or
mandates for
risk reduction
Affordability
addressed?
National Flood
Insurance
Program (NFIP)
Flood only ~50% in 100-
year
floodplains,
much lower
outside them;
mandatory for
those with
loans in 100-
year
floodplain
Premiums,
debt from
Treasury,
recent
reinsurance
purchase
NFIP has
borrowing
authority
from the
Treasury;
without
borrowing it
can pay
claims up to
a 1/16.7 year
all flood
perils
occurrence
exceedance
probability
(as of 2014)b
Lower
premiums for
elevation and
some
community
measures;
regulations on
development
in 100-year
floodplains
Historically
lower rates
for older
homes
(being
phased out);
continued
discounts for
certain
properties
that face
increased
risk
California
Earthquake
Authority
(CEA)
Residential
earthquake
only
~10%;
mandatory
offer, no
mandatory
purchase
Premiums,
reinsurance,
insurer
contributions
and
assessments,
debt,
accumulated
capital
1/250 year
all perils
occurrence
exceedance
probability
(as of 2014)b
Premium
discounts for
seismic
retrofits on
older homes
and for
mobile homes
reinforced by
earthquake-
resistant
bracing
system
No
Florida Citizens
Property
Insurance
Corporation
Homeowners
policies or
wind-only
policies
Wind
coverage
required by
lenders
Premiums,
reinsurance,
catastrophe
bonds, state
Hurricane
Catastrophe
Fund,
postloss
assessments
on Citizens
and non-
Citizens
policyholders
1/100 year
all perils
occurrence
exceedance
probability
(as of
2013)b; as of
2016, no
postevent
assessments
for 1/100
year event
Premium
discounts for
verified wind-
resistant
features (e.g.,
strengthening
roof-to-wall
connections,
protecting
window
openings and
doors)
No
Resources for the Future Kousky and Kunreuther
9
Florida
Hurricane
Catastrophe
Fund
Mandatory
reinsurance
to companies
writing
policies in
Florida
Mandatory
participation
for authorized
property
insurers
Premiums,
reinsurance,
investment
income,
revenue
bonds
N/Ad
None No
Terrorism Risk
Insurance Act
(TRIA)
Reinsurance
for terrorism
losses in
United States
~60%;
mandatory
offer; no
mandatory
purchase
Federal
funds,
surcharges on
commercially
insured
policyholders
Claims
payments
guaranteed
by US
government
for industry
losses up to
$100 billion
None No, but
evidence
rates are a
small
percentage
of total
premiums
Flood Re Reinsurance
for flood
losses in
United
Kingdom
~95% take-up
for primary
coverage;
flood coverage
required for
mortgage and
included in
standard
homeowners
policies
Premiums,
levy on
insurers, ad
hoc payments
from insurers
1/200 year
eventc
None Temporary
premium
subsidies for
high-risk
properties
(shifting to
risk-based
pricing over
20–25 years)
All hazards
Insurance
All natural
hazards
including
flood,
earthquake,
fire, wind,
and hail
Generally
mandatory,
varies by
country
Varies by
country but
may include
premiums,
policy
surcharges,
reinsurance,
government
funds, etc.
Varies by
insurer and
country
Varies and
depends on
whether
pricing is
risk-based
Varies;
uniform
rates or flat
fees in some
countries
(e.g., Spain,
France, New
Zealand)
a This column gives the maximum event, expressed by return interval, for which the program could pay claims.
b These estimates are taken from NFIP (2015).
c Taken from Surminski (2017).
d The FHCF has a statutory limit of up to $17 billion, but this will provide different claims-paying capacity to the
150-plus participating insurers. Without detailed data from each insurance company, it is difficult to determine when
FHCF would exhaust its $17 billion. See Medders and Nicholoson (2017) for more discussion.
Table 1 contains useful lessons. Well-functioning disaster insurance markets for extreme
events—earthquakes, floods, hurricanes, terrorism—require some type of government
intervention. These public sector policies and programs are diverse, often driven in part by a
country’s culture and historical experience. While there does not appear to be one best fits-all
design, some approaches have proven less effective or have more perverse or unintended
consequences than others, as discussed in the following sections. Take-up of disaster insurance is
generally low absent a mandate from either the government or lenders. Very few insurance
Resources for the Future Kousky and Kunreuther
10
programs charge premiums that fully reflect the risk at a property level because the likelihood
and consequences of future losses are difficult to estimate, the premiums are averaged over large
areas, and/or the premiums are cross-subsidized or discounted for various reasons, such as
affordability concerns or regulatory restrictions on premiums for high-risk properties. Some
actively engage in risk communication and hazard mitigation, while others do less of these
activities. There is large variation in claims paying ability and the financing used by these
programs. At one extreme is the CEA, which has access to capital to pay claims up to a 1-in-250-
year event. At the other extreme is the NFIP, which is already in debt and so has very little
capital to pay claims, although it would presumably have borrowing authority from the Treasury
(it must be increased by Congress).
4. Roles of the Public and Private Sectors
Workshop participants and the authors of the papers in this Special Issue have identified
challenges to insuring against extreme events and offered concrete recommendations for dealing
with them in three related areas: (1) risk communication, (2) risk reduction, and (3) risk transfer.
We consider the comparative advantages of the public and private sectors and how they can
work together effectively in each of these areas.
4.1. Risk Communication
As discussed above, decisionmakers often poorly evaluate low-probability risks,
exhibiting systematic biases. It is thus not surprising that a major challenge is communicating
risk so that those who are exposed to the risk are aware and understand the potential
consequences. Workshop participants agreed that promoting risk understanding and acceptance
of technical risk information was essential and had to be undertaken by both private and public
sectors.
Insurance can play an important role in communicating risk if premiums are transparent
and if the pricing reflects the risk. A premium with a hidden or unexplained cross-subsidy or
discount, however, may lead people to believe they are safer than they actually are. There was a
consensus that property-level risk based pricing would be an improvement over current practices.
Many insurance programs, particularly those where the public sector plays a key role,
have risk communication beyond pricing as an explicit objective. For instance, the website of the
Resources for the Future Kousky and Kunreuther
11
CEA details earthquake risk for different counties in California and provides information on
what structures may be suited for retrofits.3 The NFIP provides flood risk information through
mailings and online. It also provides flood hazard maps to all participating communities. Since
the maps were created to implement the program and not simply as the best risk communication
tool for local governments, however, they might actually mislead residents in flood-prone areas.
Specifically, the maps define a high risk area as one where the chance of flooding has an annual
probability greater than 1 in 100. This line may lead property owners to believe that if they live
outside this zone they are safe and all areas within it are equally risky (Kousky 2017).
Beyond the insurance programs, there is also a role for public sector information
provision. In the U.S., the federal government already has risk communication efforts, such as
Flood Smart and the Federal Alliance for Safe Homes (FLASH). Flood Smart is the NFIP’s
primary platform for communicating flood risk and insurance information to the public. The
campaign runs messages on radio, television, and in print, but its primary outreach tool is
FloodSmart.gov, a website with detailed information about flood hazards, maps, and insurance
coverage. FLASH, a collaborative disaster safety education organization comprising more than
100 state and federal emergency management agencies, (re)insurers, business, nonprofits, and
others, creates disaster education programs, publishes home safety and insurance guidebooks,
and offers a wide range of resources, including videos, peril maps, and toolkits for homeowners
to make safety improvements on their own. There are also myriad communication activities at
the state and local level. For example, New York City has created the website
http://www.floodhelpny.org to give residents information on flood risk and mitigation options.
Workshop participants agreed that a critical time for information provision is during the
early stage in the home purchase decision. Mandatory hazard disclosure laws at the time property
is sold could ensure that all property owners are informed about disaster risks. Nationally,
potential purchasers of property in 100-year floodplains are required to be informed by the lender
at the time of sale that the property is in a mapped high risk area. Some states have more
expansive disclosure requirements. California, for instance, requires sellers to disclose whether a
property is in a designated flood, wildfire, or seismic area. A study of California’s 1998 flood
hazard disclosure law found that it did have an effect: floodplain properties sold for 4-plus
percent less after the law than before it (Troy and Romm 2004). For such policies to be effective,
3 See https://www.earthquakeauthority.com/earthquake-risk-preparedness.
Resources for the Future Kousky and Kunreuther
12
disclosure should be made early in the sale process—information disclosed just before the sale is
finalized is unlikely to be heeded, given the buyer’s psychological commitment to the property
(Palm 1981)—and the requirement should be strongly enforced (Godschalk et al. 2000).
Currently, it is rare for property owners to receive any information on how the risk may
change over the coming years or decades. This is most relevant for coastal flood risks, where
sea-level rise and changing storm patterns are likely increasing. The Technical Mapping
Advisory Council, which reviews flood hazard maps for the Federal Emergency Management
Agency (FEMA), has published “Future Conditions,” a report that highlights the importance of
considering how climate change may alter flood risk in the coming years (Technical Mapping
Advisory Council 2015). The New York City website mentioned earlier is also actively
communicating information about changing risk.
The private sector—both for-profit companies and not-for-profit groups focused on risk
management—can complement public sector efforts to improve risk communication. For
example, Temblor is a website (http://temblor.net) and mobile app that provides property-
specific seismic risk data in easy-to-understand formats while also providing information on risk
reduction activities and earthquake insurance (directing users to an agent for a property-specific
insurance quote). As another example, the mission of the Institute for Building and Home Safety,
a nonprofit supported by the insurance and reinsurance industry, is “to conduct objective,
scientific research to identify and promote effective actions that strengthen homes, businesses,
and communities against natural disasters and other causes of loss.”4
Recommendations
Make available to property owners and communities understandable, structure-specific
risk information, perhaps presented as a risk score or metric for each hazard. Disseminate
the information via multiple platforms, including websites, phone-based apps, and print
mailings. Production of such tools will likely require collaboration between the public
and private sectors.
Give homes in risk-prone areas seals of approval or scores based on their susceptibility to
different hazards. This would let potential buyers see how well the structure is protected
against future disasters. Property inspections should be conducted by independent,
4 https://disastersafety.org/about/
Resources for the Future Kousky and Kunreuther
13
certified individuals and supported by real estate agents, financial institutions, and
communities.
Develop community maps delineating current and future flood risk.
Coordinate the various government risk communication efforts across agencies and scales
of government, to promote synergies and avoid duplication of effort.
4.2. Risk Reduction
In the United States it appears that cost-effective mitigation measures are not undertaken
by many property owners and renters in hazard-prone areas. For instance, only an estimated 10
percent of earthquake- and flood-prone households have implemented mitigation measures.
Reasons for the lack of mitigation may be that property owners underestimate both the risk of a
future disaster and the potential benefits of mitigation, and others may not have access to capital
to cover the up-front costs (Kunreuther et al. 2013).
Misaligned incentives for risk reduction may suppress mitigation investments. Many
workshop participants observed that local governments reap the benefits of economic
development through increased tax revenue but bear none of the costs when a disaster occurs.
Hence they lack economic incentives to adopt strong building codes or limit development in the
highest-risk areas. Similarly, there was concern that federal assistance under the Stafford Act
may make local governments less interested in mitigation: federal disaster relief covers at least
75 percent of the cost to replace or repair damaged public buildings and infrastructure. Empirical
research on any incentive effects from federal assistance to local goverments is, however,
lacking.
Insurance and mitigation may be viewed as substitutes if the mitigation measures reduce
risks enough that property owners can comfortably self-insure. But for most cases, mitigation
can be more appropriately viewed as a complement to insurance if premiums reflect risk since it
lowers potential losses and hence insurance claims. This effect could motivate partnerships
involving private insurers, public or quasi-public insurers, and the government in encouraging
risk reduction.
Charging risk-based prices for insurance and providing long-term home improvement
loans for cost-effective hazard risk reduction could overcome consumers’ tendency to be myopic
and not look beyond a short time horizon. If the upfront cost of the mitigation measure is spread
over a long period, such as the term of the mortgage, and the annual insurance premium is
reduced, investing in mitigation could be advantageous to the property owner (Kunreuther 2008).
Resources for the Future Kousky and Kunreuther
14
The Small Business Administration already offers loans for disaster victims to include hazard
mitigation in their rebuilding, although very few households take advantage of this program.
FEMA has produced outreach materials to inform homeowners that if a rebuilding loan includes
the cost of elevation, it can often be cheaper to elevate several feet above the minimum due to
the much lower flood insurance premium they receive for such elevation (FEMA 2013). Xian et
al. (2017) has determined the optimal elevation based on estimates of the NFIP risk-based
insurance premium coupled with a long-term loan at different interest rates and payback periods.
Several insurance programs and companies offer premium reductions for certain
mitigation measures. For example, the NFIP offers substantially lower rates when homes are
elevated above the base flood elevation (see Kousky et al. 2016). The CEA now offers premium
discounts up to 20 percent for new, safer construction (Marshall 2017). In addition, some
states—including Alabama, California, Florida, Louisiana, Maryland, Mississippi, New York,
South Carolina, and Texas—require companies to offer premium discounts for certain hazard
mitigation measures, or they have state insurance programs that offer such discounts
(Multihazard Mitigation Council 2015; OMB 2016). It is unclear, however, whether these
programs have encouraged homeowners to invest in new hazard mitigation measures or simply
rewarded homeowners who have already done so.
Florida provides a cautionary tale on the potential problems with state-mandated
premium discounts. In 2003, it required insurance companies to provide discounts for certain
mitigation options. Unfortunately, the way the state enacted this requirement was such that it
forced companies to provide premium discounts that were unjustified and not matched to claims
payments (for more detail, see Medders and Nicholson 2017). Moreover, accountability and
oversight were insufficient. In the end, more than half of inspected homes obtained premium
credits without undertaking any mitigation (Medders et al. 2014), and these financially unsound
credits may have prompted insurers to leave the state or damaged firms’ financial positions
(Florida Commission on Hurricane Loss Methodology 2010).
Governmental disaster insurance programs can have unique opportunities to promote or
fund hazard mitigation. Risk reduction is an explicit goal of the NFIP, with FEMA offering
grants to encourage homeowners to mitigate properties that flood repeatedly. Reducing losses
through cost-effective mitigation is critical to reducing future claims, since unlike a private
insurer, the program cannot refuse coverage to any property in a participating community. The
CEA has two grant programs for mitigating properties at greatest risk of earthquake damage.
One is for risky properties that have held a CEA policy for three or more years. A qualifying
Resources for the Future Kousky and Kunreuther
15
homeowner who braces and bolts the insured structure receives a $3,000 grant to offset the costs,
as well as a 20 percent reduction on CEA premiums.5
The federal government has small amounts of grant dollars available for supporting risk
reduction projects during the pre-disaster period. One example is FEMA’s Pre-Disaster
Mitigation grant program; the Corps of Engineers also invests in risk reduction prior to disaster
events. For flooding, a recent analysis of federal risk reduction spending—which includes
programs in FEMA, the Corps, and the Department of Housing and Urban Development, among
a few others—finds that more than 90 percent of the funding is tied to a presidential disaster
declaration post-disaster and limited to the affected community; very little funding is available
pre-disaster (Kousky and Shabman 2017).
Local governments can reduce risk through well-enforced building codes and land-use
regulations. The NFIP has a unique lever to influence these decisions, since flood insurance is
available to a community only if the local government adopts floodplain regulations, including
that all new construction in the 100-year floodplain to be built at or above the base flood
elevation. Beyond this minimum requirement, the NFIP tries to encourage voluntary risk
reduction by local governments through the Community Rating System. If a community takes
steps to lower its flood risk, all property owners receive discounted flood insurance premiums.
Flood Re, the flood insurance program in the United Kingdom, has not yet taken steps to
encourage investment in risk reduction measures, a potentially missed opportunity (Surminsky
2017).
Recommendations
Expand federal and state funding for predisaster hazard mitigation, particularly for the
riskiest properties, such as those that experience repetitive flooding.
Examine how insurance premium reductions could create incentives for new mitigation
investments, perhaps in combination with low-interest loans.
Promote funding and financing that are underutilized, such as postdisaster Small Business
Administration loans for hazard mitigation.
Convene an expert group to investigate better ways of incorporating the full costs of
insuring a structure over its life into initial decisions of where and how to build.
5 For more information, see the program website: https://www.ceabracebolt.com/.
Resources for the Future Kousky and Kunreuther
16
4.3. Risk Transfer
Risk transfer for disasters can be costly, for the reasons discussed in Section 2. We begin
with the proposition that the private market should insure and reinsure disaster risks to the extent
it can, recognizing that the public sector will need to play three complementary roles to facilitate
equitable and efficient transfer of risk: (1) making disaster insurance affordable for low-income
households, (2) providing public reinsurance against catastrophic losses that are beyond the
capabilities of reinsurers and other risk transfer instruments, and (3) mandating insurance
coverage, if widespread take-up rates is a policy objective.6
The cost of disaster insurance is often high, and the cost burden of flood insurance
premiums on low-income households has emerged as a policy concern (National Research
Council 2015). If disaster coverage is deemed critical for such households, help should come
through an explicit federal assistance program rather than through a cross-subsidy that distorts
risk signals and is viewed as inequitable by those who are charged higher premiums to cover the
cross-subsidy.
Such pricing distortions have been used in several of the programs evaluated, including
the NFIP and Florida Citizens. In the NFIP, lower premiums were provided, not based on
household income or wealth, but on the age of the structure: older homes were given discounted
rates, which rendered the program unable to finance catastrophic losses (Kousky 2017). Due to
legislation in 2012 and 2014, these are now being phased out. Beginning in 2005, Florida
Citizens Property Insurance Corporation, a nonprofit, tax-exempt government entity, effectively
froze premiums that were highly subsidized; the private sector could not compete, and Citizens
became the largest insurance provider in the state. Private insurers’ surplus plummeted, and
many left the state (Medders and Nicholoson 2017). The CEA, on the other hand, has not
discounted premiums in any manner that would threaten the financial standing of the program,
but as a result, take-up rates are very low, hovering around 10 percent (Marshall 2017). To the
extent higher premiums deter the purchase of insurance, risk based pricing may discourage some
property owners from purchasing coverage voluntarily.
6 The public sector must also, of course, continue to play a role in consumer protection regulations. In the United
States, state-level regulators protect the consumer by requiring that insurers have sufficient capital and surplus to
pay claims following a severe disaster. These capital requirements are designed to minimize insolvency risk.
Regulators also approve insurance pricing to protect the consumer from unfair premiums.
Resources for the Future Kousky and Kunreuther
17
For disaster risks with a very long tail of the loss distribution and risks that are difficult to
model, states and the federal government can cover extreme losses above a large threshold. This
is the case today with terrorism coverage. After 9/11, as discussed above, insurers refused to
offer terrorism coverage or did so at exorbitant prices, and in response Congress passed the
Terrorism Risk Insurance Act (TRIA), which requires primary insurance companies to offer
terrorism coverage in exchange for the federal government’s covering a portion of catastrophic
losses. This is not a complete federal backstop of losses above some threshold, since the federal
government can recoup its payments ex post through commercial policyholder assessments. The
arrangement has made terrorism coverage widely available and affordable for commercial
properties (Michel-Kerjan and Kunreuther 2017). In contrast, a federal backstop was never
explicitly guaranteed for the NFIP, and now the program carries a debt that it will never be able
to repay.
The Florida Hurricane Catastrophe Fund (FHCF), an example of US state reinsurance,
was created in 1993 to provide more capacity for private insurers to offer coverage in the state.
Florida Citizens and private insurers are required to purchase protection from the FHCF. Surplus
in the FHCF can accumulate tax-free. In the event of a large disaster, it can issue bonds to pay
claims and then repay this debt through assessments on policyholders in most lines of property
casualty insurance. Although FHCF pricing is required to be actuarial, it is still below the costs
of private sector reinsurance, since it is tax-exempt and does not need to include the risk and
profit loads or marketing costs and expenses that private reinsurers incorporate in their pricing
(Medders and Nicholoson 2017).
In other countries, the government covers losses above a high threshold, either as a
backstop or as reinsurance. For example, in Spain, where “extraordinary risk” insurance is a
mandatory component of property, life, and personal accident policies, the state-run Consorcio
de Comepensacion de Seguros is responsible for insuring against natural hazards and social-
political perils and operates with an unlimited government guarantee. In France, insurers are
required to provide natural catastrophe coverage as part of “physical damage” contracts and may
transfer those risks to the state-owned reinsurer, Caisse Centrale de Réassurance, at a subsidized
cost. It is secured by a government backstop, but payouts are contingent on an official
declaration of a natural catastrophe. In New Zealand, the government’s Earthquake Commission
provides natural disaster protection automatically to homeowners who insure against fire. The
program administers and pays claims from the state’s natural disaster fund, which is backed by
an unlimited government guarantee.
Resources for the Future Kousky and Kunreuther
18
Another approach to helping the private sector manage catastrophic losses is to set up a
separate program for the riskiest properties, as Flood Re has done in the United Kingdom
(Surminski 2017). Insurers can cede the riskiest properties to the Flood Re pool at a discounted
price, paid for by a levy on all insurers according to their home insurance market share. Because
the private insurance company’s liability for these properties is reduced, it can write coverage for
the remaining properties in flood-prone areas of the country. In this model, as with the Florida
programs, the cost is distributed over a wider base of policyholders, not just those who have
suffered a loss.
Similarly, residual market mechanisms in the United States were designed to provide
coverage for those unable to find a policy in the voluntary market. The Insurance Information
Institute has found that exposure in beach and wind pools has been growing, suggesting these are
not markets of last resort but often primary markets for wind coverage in hurricane-prone areas
(Hartwig and Wilkinson 2014). In Florida, Citizens explicitly chose to offer lower-cost policies
and became the largest insurer in the state (Medders and Nicholoson 2017). More recently, it has
been shedding policies to better manage exposure.
The benefits of widespread take-up of disaster coverage include reduced rebuilding time,
faster resumption of normal economic activity, the decreased need for federal relief, and
enhanced resilience. If increasing take-up is a policy goal, it may be necessary to require
homeowners to purchase disaster insurance. Substantial evidence indicates that households do
not voluntarily insure against earthquakes and floods, and lenders in the United States do not
require insurance against earthquake risk and only require flood when mandated by law. As
pointed out in Section 2, many homeowners in flood-prone areas who are required by the NFIP
to have coverage are uninsured today. Bundling flood and earthquake insurance into the
homeowners coverage that banks require as a condition for a mortgage would dramatically
increase take-up rates (Kunreuther 2017). Flood coverage is much more widespread in the
United Kingdom, where it is a part of homeowners insurance and required by lenders (Surminski
2017).
Recommendations
Develop a federal assistance program for low-income households so that they can afford
disaster coverage. This should be paid for by general tax dollars and not through cross-
subsidies in insurance pricing.
Determine whether government backstops or reinsurance against catastrophic losses not
covered by private reinsurance or other risk transfer mechanisms, such as catastrophe
Resources for the Future Kousky and Kunreuther
19
bonds, could strengthen and improve the functioning of the primary insurance market.
The study should identify the level at which government support is made available and
define the structure and process for receiving that support.
Analyze when and how risky properties could be ceded to a government residual market
and how those risks would be financed.
Promote better relationships between state regulators and insurance companies on
disaster lines and innovation in product offerings for disaster lines.
5. A Proposed Program
We conclude with a proposal for a comprehensive program that capitalizes on the
strengths of the public and private sectors while requiring property owners to absorb some of the
risk associated with disasters so that they have incentive to undertake mitigation measures. We
base this proposal on several guiding principles:
Premiums should reflect risk, with information on the hazards clearly communicated to
prospective policyholders.
Cost-effective mitigation measures should be encouraged through insurance premium
reductions and long-term loans that spread the upfront costs over time.
Affordability of insurance for low-income residents should be addressed through public
sector assistance, not subsidized insurance premiums.
Catastrophic losses that cannot be covered by private reinsurance or other risk transfer
instruments should be backstopped by the public sector.
The public sector should provide incentives, information, and some funding for
investments in risk reduction.
All natural disasters should be covered in a single private insurance policy to reduce the
variance in losses and satisfy homeowners’ desire to be covered against all hazards.
We propose a system of layering of disaster losses as shown in Figure 1. The blue boxes
show schematically how the lowers would function; the size is not necessarily proportional to the
amount of coverage in each layer. This layering idea is not new, and it has many appealing
features, such as providing more appropriate economic incentives, enhancing the involvement of
the private sector, and better meeting the financial needs of those at risk from future disasters
Resources for the Future Kousky and Kunreuther
20
(e.g., Litan 2006). We note, however, that the layers of risk transfer need to be supported by
public and private sector activity centered on risk communication and risk reduction.
Figure 1. Layering Disaster Losses
The property owner would be responsible for the first layer of losses through a
deductible, which creates an incentive to reduce future losses by undertaking mitigation
measures while reducing moral hazard. Property owners could choose higher deductibles that
would lower their insurance premiums if they were able to self-insure a higher portion of losses.
Higher deductibles might be supported by tax-exempt disaster savings accounts, which property
owners could tap to cover losses.
The second layer of losses would be transferred to an insurer through a homeowners
policy that included all natural perils, with premiums reflecting risk. Banks and financial
institutions would require this coverage as a condition for a mortgage.7 An all-hazards policy has
numerous benefits for the property owner. As the National Association of Insurance
Commissioners has observed, having to purchase multiple policies is inefficient and confusing,
and it delays claims settlements because the cause of the losses from hurricanes—wind or
water—must first be determined (NAIC 2009). Following Hurricane Katrina, insurers spent
considerable time and money fighting lawsuits from policyholders who contended that damage
to their homes was due to wind, in which case the standard homeowners policy would cover the
losses, rather than inundation. An all-hazards policy would avoid this problem (Kunreuther
7 The NAIC has suggested an insurance mandate in highest-risk zones for property owners with federally backed
mortgages. This is similar to the current flood insurance requirement by the NFIP but includes all natural disasters
(NAIC 2009).
Resources for the Future Kousky and Kunreuther
21
2017). Financial assistance for low-income homeowners would be provided by the government
in form of tax credits or vouchers based on specified criteria for household income or house
value. The tax credit or voucher could also cover the cost of a loan to enable the property owner
to invest in cost-effective mitigation measures (Kousky and Kunreuther 2014).
The next layer of losses would be covered through private reinsurance or other forms of
risk transfer, such as catastrophe bonds purchased by the insurance company. The highest layer
of losses would be covered by the public sector. Although neither has been tested yet with a
catastrophic loss event, the experience of both Flood Re and TRIA suggests that if the
government can cover the riskiest properties, or cover catastrophic losses, the private market
could profitably offer disaster insurance for the remaining losses above the deductible. Coverage
at the state or federal level could entail ex ante premiums or ex post assessments to recoup some
or all of the public sector expenditures. The attachment point for state or federal payments would
need to be carefully determined. It should not be too low, since the private market should be
encouraged to bear as much risk as possible, but it also needs to make private firms feel
comfortable about bearing the remaining risk. This would require a detailed market evaluation.
As shown in Figure 1, both the public and private sectors would need to also engage in
supporting activities, such as risk communication, adoption and enforcement of building codes,
and limitations on future development through restrictive zoning and other land-use regulations.
6. Conclusion and Suggestions for Future Research
Disasters are challenging to insure, and the price of coverage is often high. In response,
policymakers around the world have devised a variety of quasi-public and fully public disaster
insurance schemes. In this paper, we have teased out some guidelines for how the public and
private sector can work together to move toward a system in which comprehensive and
affordable coverage is provided for those at risk. That said, it is clear that more detailed research
and investigation are needed to inform reform efforts of the current systems. In addition, new
policies may be required. We highlight several in this concluding section.
6.1. Finer-Scale Risk Communication
Participants in the workshop agreed that better information on the specific risks at a given
property could encourage investments in risk reduction and promote the purchase and retention
of disaster insurance. Because prices send a power signal about risk, fine-grained insurance
pricing could also improve risk understanding. Communication efforts presuppose better
Resources for the Future Kousky and Kunreuther
22
mapping and modeling. For flood, this means investment in more data and application of
advanced modeling techniques. Where data are not available or cost-effective, lower-cost
options, need to be explored. All approaches require funding, however, and thus Congress should
appropriate funds for continued improvements in hazard mapping. Once better maps are
available, researchers in risk communication should investigate the most effective ways of
communicating the results to residents.
6.2. Higher Take-Up of Insurance
Various factors can explain why so many people at risk have no disaster insurance and
fail to invest in hazard mitigation: homeowners may lack information on the risk, dismiss it as
negligible, be myopic, lack the necessary capital, regard the transaction costs as too high, or
respond to perverse incentives. Field research is needed to untangle which contributing factors
are most important and whether policy changes could help counter the tendency to underinvest in
insurance and mitigation.
One of the biggest challenges is to convince homeowners that not incurring losses from a
disaster does not make the purchase of insurance a waste of money or justify canceling the
policy. It is extremely difficult to get the message across that people should celebrate not having
collected on their insurance policy this year, and that a disaster could still occur next year. Future
empirical and experimental research could determine the effectiveness of rebates for those who
have not suffered a loss coupled with messages such as “the best return on an insurance policy is
no return at all.”
6.3. Insurance for Low-Income Households
One part of closing the disaster insurance gap is making sure that all those living in risky
areas can afford disaster insurance. Workshop participants suggested that instead of using cross-
subsidies in pricing, Congress create an explicit program to provide assistance to households in
need, coupled with a mitigation program. This approach would reduce both the losses incurred
by low-income residents and the costs of risk transfer. There remain several unanswered
questions, however: how much assistance would be needed, what the scope of the program
would be, and what types of mitigation could be offered, among others. A good example of the
type of studies needed was recently completed by RAND (Dixon et al. 2017).
Resources for the Future Kousky and Kunreuther
23
6.4. New Policy Models
This paper has proposed several roles for the public sector in improving insurance of
disasters, and an even broader array of policies have been proposed in scholarly work on this
topic. More research is needed on comparing and evaluating these alternatives. For example,
would a residual pool model or a federal reinsurance model be preferable—and on what
grounds—to the current federal program of direct writing of flood coverage? Such work first
requires developing a consensus on the public policy goals that should guide program
development.
Resources for the Future Kousky and Kunreuther
24
References
Adams, I. 2015. Earthquake risks are rising, but take-up isn’t. Insurance Journal, March 11.
Atreya, A., S. Hanger, H. Kunreuther, J. Linnerooth-Bayer, and E. Michel-Kerjan. 2015. A
comparison of residential flood insruance markets in 25 countries. Wharton Risk
Management and Decision Processes Center, University of Pennsylvania, Philadelphia,
June 28.
Benson, C., and E. J. Clay. 2004. Understanding the economic and financial impacts of natural
disasters. Washington, DC: The World Bank,
Browne, M. J., and R. E. Hoyt. 2000. The demand for flood insurance: Empirical evidence.
Journal of Risk and Uncertainty 20(3): 291–306.
Consorcio de Compensacion de Seguros (CCS). 2008. Natural catastrophes insurance cover, a
diversity of systems. Consorcio de Compensacion de Seguros, Madrid.
Dixon, L., N. Clancy, S. A. Seabury, and A. Overton. 2006. The National Flood Insurance
Program’s market penetration rate: Estimates and policy implications. Santa Monica,
California: RAND Corporation, February.
Dixon, L., N. Clancy, B. M. Miller, S Hoegberg, M. M. Lewis, B. Bener, S. Ebinger, M. Hodges,
G. M. Syck, C. Nagy, and S. R. Choquette. 2017. The Cost and Affordability of Flood
Insurance in New York City: Economic Impacts of Rising Premiums and Policy Options
for One- to Four-Family Homes. RAND Corporation, Santa Monica, CA.
Federal Emergency Management Agency (FEMA). 2013. Can we afford to rebuild higher? Can
we afford not to? Washington, DC.
Florida Commission on Hurricane Loss Methodology. 2010. Windstorm mitigation discounts
report to the governor, the cabinet, the president of the Senate, and the speaker of the
House of Representatives. Tallahassee, February.
Godschalk, D. R., R. Norton, C. Richardson, and D. Salvesen. 2000. Avoiding coastal hazard
areas: Best state mitigation practices. Environmental Geosciences 7(1): 13–22.
Hartwig, R. P., and C. Wilkinson. 2014. Residual market property plans: From markets of last
resort to markets of first choice. Insurance Information Institute, September.
Hogarth, R. M., and H. Kunreuther. 1985. Ambiguity and insurance decisions. American
Economic Review 75(2): 386–390.
Resources for the Future Kousky and Kunreuther
25
Holmes, T. P., R. J. Huggett Jr., and A. L. Westerling. 2008. Statistical analysis of large
wildfires. In T. P. Holmes, J. P. Prestemon and K. L. Abt (eds.), The economics of forest
disturbances: Wildfires, storms, and invasive species. Springer Science, 59–77.
Intergovernmental Panel on Climate Change (IPCC). 2012. Managing the risks of extreme events
and disasters to advance climate change adaptation. Cambridge: Cambridge University
Press.
Jaffee, D. M., and T. Russell. 2003. Market under stress: The case of extreme event insurance. In
R. Arnott, B. Greenwald, R. Kanbur and B. Nalebuff (eds.), Economics for an Imperfect
World: Essays in Honor of Joseph E. Stiglitz. Cambridge, MA: MIT Press.
Kahneman, D. 2011. Thinking, Fast and Slow. New York: Farrar, Straus, and Giroux.
Klein, R. 2008. Catastrophe risk and the regulation of property insurance. Center for RMI
Research, Georgia State University, Atlanta.
Klein, R., and S. Wang. 2007. Catastrophe risk financing in the US and EU: A comparative
analysis of alternative regulatory approaches. SCOR-JRI Conference on Insurance,
Reinsurance, and Capital Market Transformations, Paris.
Knowles, S. G., and H. C. Kunreuther. 2014. Troubled waters: The National Flood Insurance
Program in historical perspective. Journal of Policy History 26(3): 327–353.
Kousky, C. 2014. Informing climate adaptation: A review of the economic costs of natural
disasters. Energy Economics 46: 576–592.
———. 2016. Disasters as learning experiences or disasters as policy opportunities? Examining
flood insurance purchases after hurricanes. Risk Analysis. doi: 10.1111/risa.12646.
———. 2017. Financing flood losses: A discussion of the National Flood Insurance Program.
Discussion Paper 17-03. Washington, DC: Resources for the Future, February.
Kousky, C., and R. Cooke. 2012. Explaining the failure to insure catastrophic risks. Geneva
Papers on Risk and Insurance: Issues and Practice 37(2): 206–227.
Kousky, C., and H. Kunreuther. 2014. Addressing affordability in the National Flood Insurance
Program. Journal of Extreme Events 1(1). doi: 10.1142/S2345737614500018.
Kousky, C., B. Lingle, and L. Shabman. 2016. NFIP premiums for single-family residential
properties: Today and tomorrow. Washington, DC: Resources for the Future,
Resources for the Future Kousky and Kunreuther
26
Kousky, C., E. Michel-Kerjan, and P. Raschky. 2014. Does Federal Disaster Assistance Crowd
Out Private Demand for Insurance? Risk Management and Decision Processes Center,
Wharton School, University of Pennsylvania, Philadelphia.
Kousky, C., and L. Shabman. 2017. Federal funding for flood risk reduction in the US: Pre- or
post-disaster? Water Economics and Policy 3(1): 1771001.
Kunreuther, H. 2008. Reducing losses from catastrophic risks through long-term insurance and
mitigation. Social Research 75: 905–932.
———. 2017. All-hazards homeowners insurance: Challenges and opportunities. Discussion
Paper 17-08. Washington, DC: Resources for the Future, February.
Kunreuther, H., and R. M. Hogarth. 1992. How does ambiguity affect insurance decisions? In G.
Dionne (ed.), Contributions to insurance economics. Dordrecht: Springer, 307–324.
Kunreuther, H., and E. Michel-Kerjan. 2011. At War with the Weather (Cambridge: MIT Press)
Kunreuther, H., et al. 1978. Disaster insurance protection: Public policy lessons. New York:
Wiley.
Kunreuther, H., M. V. Pauly and S. McMorrow. 2013. Insurance and behavioral economics:
Improving decisions in the most misunderstood industry. Cambridge: Cambridge
University Press.
Lindsay, B. R., and J. Murray. 2014. Supplemental appropriations for disaster assistance:
Summary data and analysis. Washington, DC: Congressional Reserach Service, October
1.
Litan, R. E. 2006. Sharing and reducing the financial risks of future “mega-catastrophes.” Issues
in Economic Policy. Washington, DC: Brookings Institution, March.
Malamud, B. D., and D. L. Turcotte. 2006. The applicability of power-law frequency statistics to
floods. Journal of Hydrology 322: 168–180.
Marshall, D. 2017. The California Earthquake Authority. Discussion Paper 17-05. Washington,
DC: Resources for the Future,
McAneney, J., D. McAneney, R. Musulin, G. Walker, and R. Crompton. 2016. Government-
sponsored natural disaster insurance pools: A view from down-under. International
Journal of Disaster Risk Reduction 15: 1–9.
Resources for the Future Kousky and Kunreuther
27
McChristian, L. 2012. Hurricane Andrew and insurance: The enduring impact of an historic
storm. Insurance Information Institute, Tampa, Florida.
Medders, L. A., K. A. McCullough, and V. Jager. 2011. A tale of two regions: Natural
catastrophe insurance and regulation in the United States and the European Union.
Journal of Insurance Regulation 30(Winter): 171–196.
Medders, L. A., and J. E. Nicholoson. 2017. Florida’s state wind pools. Discussion Paper 17-07.
Washington, DC: Resources for the Future, February.
Medders, L. A., C. M. Nyce, and J. Bradley Karl. 2014. Market implications of public policy
interventions: The case of Florida's property insurance market. Risk Management and
Insurance Review 17(2): 183–214.
Michel-Kerjan, E., and H. Kunreuther. 2017. A successful (yet somewhat untested) case of
disaster financing: Terrorism insurance under TRIA, 2002–2020. Washington, DC:
Resources for the Future, February.
Michel-Kerjan, E., S. Lemoyne de Forges, and H. Kunreuther. 2012. Policy tenure under the
U.S. National Flood Insurance Program (NFIP). Risk Analysis 32(4): 644–658.
Multihazard Mitigation Council. 2015. Developing pre-disaster resilience based on public and
private incentivization. Washington, DC: National Institute of Building and Home Safety,
Multihazard Mitigation Council in conjunction with the Council on Finance, Insurance
and Real Estate, October.
National Association of Insurance Commissioners (NAIC). 2009. Natural catastrophe risk:
Creating a comprehensive national plan. Washington, DC.
National Flood Insurance Program (NFIP). 2015. Report to Congress on reinsuring NFIP
insurance risk and options for privatizing the NFIP. Federal Emergency Management
Agency, Washington, DC.
National Research Council. 2015. Affordability of national flood insurance premiums: Report 1.
Washington, DC: National Academies Press,
New York City. 2013. PlaNYC: A stronger, more resilient New York. Mayor's Office of Long
Term Planning and Sustainability, New York City.
Office of Management and Budget (OMB). 2016. Standards and finance to support community
resilience. White House, Washington, DC, December.
Resources for the Future Kousky and Kunreuther
28
Palm, R. 1981. Real Estate Agents and Special Studies Disclosure Boulder: Institute of
Behavioral Science, University of Colorado.
Paudel, Y. 2012. A comparative study of public–private catastrophe insurance systems: Lessons
from current practices. Geneva Papers 37: 257–285.
Roy, A. D. 1952. Safety first and the holding of assets. Econometrica 20(3): 431–449.
Sander, J., J. Eichner, E. Faust, and M. Steuer. 2013. Rising variability in thunderstorm-related
U.S. losses as a reflection of changes in large-scale thunderstorm forcing. Weather,
Climate, & Society. doi:10.1175/WCAS-D-12-00023.1.
Smetters, K. A. 2004. Insuring against terrorism: the policy challenge. Brookings-Wharton
Papers on Financial Services: 139–187.
Stone, J. 1973a. A theory of capacity and the insurance of catastrophe risks (Part I) Journal of
Risk and Insurance 40: 231–43.
———. 1973b. A theory of capacity and the insurance of catastrophe risks (Part II). Journal of
Risk and Insurance 40: 339–55.
Surminski, S. 2017. Fit for purpose and fit for the future? An evaluation of the UK’s new flood
reinsurance pool. Discussion Paper 17-04. Washington, DC: Resources for the Future.
Swiss Re Institute. 2017. Sigma No. 2, Natural catastrophes and man-made disasters in 2016: A
year of widespread damages. Swiss Re, Zurich.
Talbot, T., and O. Barder. 2016. Payouts for perils: why disaster aid is broken, and how
catastrophe insurance can help fix it. Policy Paper 087. Washington, DC: Center for
Global Development, July.
Technical Mapping Advisory Council. 2015. Future conditions risk assessment and modeling.
December.
Troy and Romm. 2004. Assessing the Price Effects of Flood Hazard Disclosure under the
California Natural Hazard Disclosure Law (AB 1195). Journal of Environmental
Planning and Management 47(1): 137-162.
Tversky, A., and D. Kahneman. 1973. Availability: A heuristic for judging frequency and
probability. Cognitive Psychology 5: 207–232.
Xian, S., Lin,N. and H. Kunreuther. 2017. Optimal house elevation for reducing flood-related
losses. Journal of Hydrology 548 (May): 63–74.